svp_AR_order: AR-order selection sweep for SV(p)

In wARMASVp: Winsorized ARMA Estimation for Higher-Order Stochastic Volatility Models

AR-order selection sweep for SV(p)

Description

Convenience wrapper around svp_IC: fits svp at each p = 1, ..., pmax and returns a matrix of information criteria along with the argmin per criterion.

Usage

svp_AR_order(
  y,
  pmax = 6L,
  J = 10L,
  leverage = FALSE,
  errorType = "Gaussian",
  rho_type = "pearson",
  del = 1e-10,
  trunc_lev = TRUE,
  wDecay = FALSE,
  logNu = FALSE,
  sigvMethod = "factored",
  winsorize_eps = 0L,
  filter_method = "mixture",
  proxy = c("bayes_optimal", "u"),
  K = 7L,
  M = 1000L,
  seed = 42L,
  criteria = c("BIC_Kalman", "AIC_Kalman", "BIC_HR", "AIC_HR")
)

Arguments

y	Numeric vector. Observed returns.
pmax	Integer. Maximum AR order to consider. Default 6.
J	Integer. Winsorizing parameter passed to svp. Default 10.
leverage	Logical. Whether to estimate leverage. Default FALSE.
errorType	Character. "Gaussian", "Student-t", or "GED". Default "Gaussian".
rho_type, del, trunc_lev, wDecay, logNu, sigvMethod, winsorize_eps	Other arguments passed to svp.
filter_method	Character. Filter method for *_Kalman criteria. Default "mixture".
proxy	Character. Leverage proxy. Default "bayes_optimal" for IC consistency under Student-t leverage. See svp_IC.
K, M, seed	Filter arguments passed to filter_svp.
criteria	Character vector of criteria to compute. Default returns the four recommended criteria: c("BIC_Kalman", "AIC_Kalman", "BIC_HR", "AIC_HR"). See svp_IC for the full set of eight valid names and the rationale for each opt-in criterion.

Value

A list with components:

IC

Numeric matrix, one row per criterion, one column per candidate p in 1:pmax.

argmin

Named integer vector, one entry per criterion, giving the selected p. NA_integer_ if all entries for that criterion are NA.

fits

List of length pmax containing the fitted svp() objects (or NULL if a fit failed).

See Also

svp_IC, svp, filter_svp

Examples

set.seed(1)
y <- sim_svp(2000, phi = 0.95, sigy = 1, sigv = 0.5)$y
res <- svp_AR_order(y, pmax = 4)
res$IC
res$argmin

wARMASVp documentation built on May 15, 2026, 5:07 p.m.